The choice not to be resuscitated.

We studied the use of the do not resuscitate (DNR) order among general medical patients. During the six-month study period, of 2431 admissions there were 95 patients (3.9%) declared DNR. Of 105 deaths, 69 (66%) occurred among DNR patients. Twenty-seven percent of the DNR patients left the hospital alive. In 90% of the cases the medical intern initiated the DNR discussion, and in 60% of the cases a decision was reached in less than 24 hours. Two-thirds of the DNR orders were consented to by family members, and one-third of all DNR orders had a discernible impact on nonresuscitative care. Less aggressive therapy accounted for 60% of these restrictions, whereas the remainder involved limitations in diagnostic evaluations. When compared with age- and sex-matched general adult inpatient control subjects, DNR patients were found to have longer hospitalizations (P = .01), be more likely to reside in a nursing home (RR = 4.2), have a metastatic neoplasm (RR = 3.6), and be admitted with an abnormal mental status (RR = 6.1) or urinary incontinence (RR = 2.9). These differences remained significant when we controlled for the presence of a metastatic neoplasm. Despite a high in-hospital mortality rate (73%), DNR patients were not admitted more frequently than controls to the intensive care unit. We conclude that the DNR decision is based on clinical prognostic indexes and that once established, this order serves to modify the allocation of medical resources.